> The answers are so vague that you are not going to be able to follow up on > them. So they are unhappy with GNOME. Then what?
Then at the very least you've got some picture of what is going on and you can try and trigger discussion about why people are unhappy (or indeed happy). > Of course, maybe I'm wrong. Perhaps the average user of Linux/GNOME does > know what GNOME is, knows how to contact the GNOME team and can tell you > what version of GNOME they are using. And if they do, what is the survey There are going to be a large number of users whose viewpoint is essentially "don't care", how you measure them is hard in pretty much any situation. A truely random sample of Gnome users will be hard to get by any approach. > going to tell you? That they do or don't like GNOME? And how long they have > been using GNOME? What are we going to do with that information? Use it to work out what questions it would be interesting to ask next year ? Look at what shows up in terms of additional comments. Look at the discussion around it, drop in the odd 'Why ?' question of your own. Use it to kickstart a secondary debate on the gnome site. (And btw while they won't read LWN people will link to it and discuss it in other places too) There is a second thing here too IMHO. The questions that could be asked and fixing them are currently buried in the debate. I can't see how progress will be made on picking questions usefully until someone moves from trying to achieve consensus to picking what they think is best based upon the resposes and just doing it regardless of whether each question is considered wrong by 5% of the people in the debate. > Before any survey, you should know how you are going to use the information > so that you can be sure to ask the right questions. So I could equally have said "Why release Gnome 3.0, we know it isn't perfect and there are wrong things". Releasing it was better than stasis, it provided a learning experience that will make 3.2 much better I am sure. Doing nothing achieves nothing, doing something achieves learning. You may well not learn what you intended but you will learn something including quite possibly how to do future surveys better. I'm not saying its necessarily a great approach but it's vastly superior to people sitting around picking holes in the idea until it never happens. Right now this seems to be in blocking mode, and blocking a volunteer off to do stuff and see what happens be it code or otherwise is usually the wrong thing to do. Sure -t here is a good case for not describing it in any way that suggests its GNOME foundation endorsed or driven. Gnome grew from a comically clueless 0.1 tarball, surveys can do likewise. Alan _______________________________________________ desktop-devel-list mailing list email@example.com http://mail.gnome.org/mailman/listinfo/desktop-devel-list